Showing 1–50 of 937 results for author: Liu, G

Flow Matching Meets Biology and Life Science: A Survey

Authors: Zihao Li, Zhichen Zeng, Xiao Lin, Feihao Fang, Yanru Qu, Zhe Xu, Zhining Liu, Xuying Ning, Tianxin Wei, Ge Liu, Hanghang Tong, Jingrui He

Abstract: Over the past decade, advances in generative modeling, such as generative adversarial networks, masked autoencoders, and diffusion models, have significantly transformed biological research and discovery, enabling breakthroughs in molecule design, protein generation, drug discovery, and beyond. At the same time, biological applications have served as valuable testbeds for evaluating the capabiliti… ▽ More Over the past decade, advances in generative modeling, such as generative adversarial networks, masked autoencoders, and diffusion models, have significantly transformed biological research and discovery, enabling breakthroughs in molecule design, protein generation, drug discovery, and beyond. At the same time, biological applications have served as valuable testbeds for evaluating the capabilities of generative models. Recently, flow matching has emerged as a powerful and efficient alternative to diffusion-based generative modeling, with growing interest in its application to problems in biology and life sciences. This paper presents the first comprehensive survey of recent developments in flow matching and its applications in biological domains. We begin by systematically reviewing the foundations and variants of flow matching, and then categorize its applications into three major areas: biological sequence modeling, molecule generation and design, and peptide and protein generation. For each, we provide an in-depth review of recent progress. We also summarize commonly used datasets and software tools, and conclude with a discussion of potential future directions. The corresponding curated resources are available at https://github.com/Violet24K/Awesome-Flow-Matching-Meets-Biology. △ Less

Submitted 23 July, 2025; originally announced July 2025.

Comments: Preprint, 27 pages

GALE: Leveraging Heterogeneous Systems for Efficient Unstructured Mesh Data Analysis

Authors: Guoxi Liu, Thomas Randall, Rong Ge, Federico Iuricich

Abstract: Unstructured meshes present challenges in scientific data analysis due to irregular distribution and complex connectivity. Computing and storing connectivity information is a major bottleneck for visualization algorithms, affecting both time and memory performance. Recent task-parallel data structures address this by precomputing connectivity information at runtime while the analysis algorithm exe… ▽ More Unstructured meshes present challenges in scientific data analysis due to irregular distribution and complex connectivity. Computing and storing connectivity information is a major bottleneck for visualization algorithms, affecting both time and memory performance. Recent task-parallel data structures address this by precomputing connectivity information at runtime while the analysis algorithm executes, effectively hiding computation costs and improving performance. However, existing approaches are CPU-bound, forcing the data structure and analysis algorithm to compete for the same computational resources, limiting potential speedups. To overcome this limitation, we introduce a novel task-parallel approach optimized for heterogeneous CPU-GPU systems. Specifically, we offload the computation of mesh connectivity information to GPU threads, enabling CPU threads to focus on executing the visualization algorithm. Following this paradigm, we propose GALE (GPU-Aided Localized data structurE), the first open-source CUDA-based data structure designed for heterogeneous task parallelism. Experiments on two 20-core CPUs and an NVIDIA V100 GPU show that GALE achieves up to 2.7x speedup over state-of-the-art localized data structures while maintaining memory efficiency. △ Less

Submitted 21 July, 2025; originally announced July 2025.

Causal Language Control in Multilingual Transformers via Sparse Feature Steering

Authors: Cheng-Ting Chou, George Liu, Jessica Sun, Cole Blondin, Kevin Zhu, Vasu Sharma, Sean O'Brien

Abstract: Deterministically controlling the target generation language of large multilingual language models (LLMs) remains a fundamental challenge, particularly in zero-shot settings where neither explicit language prompts nor fine-tuning are available. In this work, we investigate whether sparse autoencoder (SAE) features, previously shown to correlate with interpretable model behaviors, can be leveraged… ▽ More Deterministically controlling the target generation language of large multilingual language models (LLMs) remains a fundamental challenge, particularly in zero-shot settings where neither explicit language prompts nor fine-tuning are available. In this work, we investigate whether sparse autoencoder (SAE) features, previously shown to correlate with interpretable model behaviors, can be leveraged to steer the generated language of LLMs during inference. Leveraging pretrained SAEs on the residual streams of Gemma-2B and Gemma-9B, we identify features whose activations differ most significantly between English and four target languages: Chinese, Japanese, Spanish, and French. By modifying just a single SAE feature at one transformer layer, we achieve controlled language shifts with up to 90\% success, as measured by FastText language classification, while preserving semantic fidelity according to LaBSE (Language-Agnostic BERT Sentence Embedding) similarity. Our analysis reveals that language steering is most effective in mid-to-late transformer layers and is amplified by specific attention heads disproportionately associated with language-sensitive SAE features. These results demonstrate the promise of sparse feature steering as a lightweight and interpretable mechanism for controllable multilingual generation. △ Less

Submitted 17 July, 2025; originally announced July 2025.

Generalist Bimanual Manipulation via Foundation Video Diffusion Models

Authors: Yao Feng, Hengkai Tan, Xinyi Mao, Guodong Liu, Shuhe Huang, Chendong Xiang, Hang Su, Jun Zhu

Abstract: Bimanual robotic manipulation, which involves the coordinated control of two robotic arms, is foundational for solving challenging tasks. Despite recent progress in general-purpose manipulation, data scarcity and embodiment heterogeneity remain serious obstacles to further scaling up in bimanual settings. In this paper, we introduce VIdeo Diffusion for Action Reasoning (VIDAR), a two-stage framewo… ▽ More Bimanual robotic manipulation, which involves the coordinated control of two robotic arms, is foundational for solving challenging tasks. Despite recent progress in general-purpose manipulation, data scarcity and embodiment heterogeneity remain serious obstacles to further scaling up in bimanual settings. In this paper, we introduce VIdeo Diffusion for Action Reasoning (VIDAR), a two-stage framework that leverages large-scale, diffusion-based video pre-training and a novel masked inverse dynamics model for action prediction. We pre-train the video diffusion model on 750K multi-view videos from three real-world bimanual robot platforms, utilizing a unified observation space that encodes robot, camera, task, and scene contexts. Our masked inverse dynamics model learns masks to extract action-relevant information from generated trajectories without requiring pixel-level labels, and the masks can effectively generalize to unseen backgrounds. Our experiments demonstrate that with only 20 minutes of human demonstrations on an unseen robot platform (only 1% of typical data requirements), VIDAR generalizes to unseen tasks and backgrounds with strong semantic understanding, surpassing state-of-the-art methods. Our findings highlight the potential of video foundation models, coupled with masked action prediction, to enable scalable and generalizable robotic manipulation in diverse real-world settings. △ Less

Submitted 17 July, 2025; originally announced July 2025.

AnyPos: Automated Task-Agnostic Actions for Bimanual Manipulation

Authors: Hengkai Tan, Yao Feng, Xinyi Mao, Shuhe Huang, Guodong Liu, Zhongkai Hao, Hang Su, Jun Zhu

Abstract: Vision-language-action (VLA) models have shown promise on task-conditioned control in complex settings such as bimanual manipulation. However, the heavy reliance on task-specific human demonstrations limits their generalization and incurs high data acquisition costs. In this work, we present a new notion of task-agnostic action paradigm that decouples action execution from task-specific conditioni… ▽ More Vision-language-action (VLA) models have shown promise on task-conditioned control in complex settings such as bimanual manipulation. However, the heavy reliance on task-specific human demonstrations limits their generalization and incurs high data acquisition costs. In this work, we present a new notion of task-agnostic action paradigm that decouples action execution from task-specific conditioning, enhancing scalability, efficiency, and cost-effectiveness. To address the data collection challenges posed by this paradigm -- such as low coverage density, behavioral redundancy, and safety risks -- we introduce ATARA (Automated Task-Agnostic Random Actions), a scalable self-supervised framework that accelerates collection by over $ 30\times $ compared to human teleoperation. To further enable effective learning from task-agnostic data, which often suffers from distribution mismatch and irrelevant trajectories, we propose AnyPos, an inverse dynamics model equipped with Arm-Decoupled Estimation and a Direction-Aware Decoder (DAD). We additionally integrate a video-conditioned action validation module to verify the feasibility of learned policies across diverse manipulation tasks. Extensive experiments show that the AnyPos-ATARA pipeline yields a 51% improvement in test accuracy and achieves 30-40% higher success rates in downstream tasks such as lifting, pick-and-place, and clicking, using replay-based video validation. Project Page: https://embodiedfoundation.github.io/vidar_anypos △ Less

Submitted 16 July, 2025; originally announced July 2025.

Scaling Up RL: Unlocking Diverse Reasoning in LLMs via Prolonged Training

Authors: Mingjie Liu, Shizhe Diao, Jian Hu, Ximing Lu, Xin Dong, Hao Zhang, Alexander Bukharin, Shaokun Zhang, Jiaqi Zeng, Makesh Narsimhan Sreedhar, Gerald Shen, David Mosallanezhad, Di Zhang, Jonas Yang, June Yang, Oleksii Kuchaiev, Guilin Liu, Zhiding Yu, Pavlo Molchanov, Yejin Choi, Jan Kautz, Yi Dong

Abstract: Recent advancements in reasoning-focused language models such as OpenAI's O1 and DeepSeek-R1 have shown that scaling test-time computation-through chain-of-thought reasoning and iterative exploration-can yield substantial improvements on complex tasks like mathematics and code generation. These breakthroughs have been driven by large-scale reinforcement learning (RL), particularly when combined wi… ▽ More Recent advancements in reasoning-focused language models such as OpenAI's O1 and DeepSeek-R1 have shown that scaling test-time computation-through chain-of-thought reasoning and iterative exploration-can yield substantial improvements on complex tasks like mathematics and code generation. These breakthroughs have been driven by large-scale reinforcement learning (RL), particularly when combined with verifiable reward signals that provide objective and grounded supervision. In this report, we investigate the effects of prolonged reinforcement learning on a small language model across a diverse set of reasoning domains. Our work identifies several key ingredients for effective training, including the use of verifiable reward tasks, enhancements to Group Relative Policy Optimization (GRPO), and practical techniques to improve training stability and generalization. We introduce controlled KL regularization, clipping ratio, and periodic reference policy resets as critical components for unlocking long-term performance gains. Our model achieves significant improvements over strong baselines, including +14.7% on math, +13.9% on coding, and +54.8% on logic puzzle tasks. To facilitate continued research, we release our model publicly. △ Less

Submitted 16 July, 2025; originally announced July 2025.

Comments: 14 pages, 7 figures

DAC: A Dynamic Attention-aware Approach for Task-Agnostic Prompt Compression

Authors: Yi Zhao, Zuchao Li, Hai Zhao, Baoyuan Qi, Guoming Liu

Abstract: Task-agnostic prompt compression leverages the redundancy in natural language to reduce computational overhead and enhance information density within prompts, especially in long-context scenarios. Existing methods predominantly rely on information entropy as the metric to compress lexical units, aiming to achieve minimal information loss. However, these approaches overlook two critical aspects: (i… ▽ More Task-agnostic prompt compression leverages the redundancy in natural language to reduce computational overhead and enhance information density within prompts, especially in long-context scenarios. Existing methods predominantly rely on information entropy as the metric to compress lexical units, aiming to achieve minimal information loss. However, these approaches overlook two critical aspects: (i) the importance of attention-critical tokens at the algorithmic level, and (ii) shifts in information entropy during the compression process. Motivated by these challenges, we propose a dynamic attention-aware approach for task-agnostic prompt compression (DAC). This approach effectively integrates entropy and attention information, dynamically sensing entropy shifts during compression to achieve fine-grained prompt compression. Extensive experiments across various domains, including LongBench, GSM8K, and BBH, show that DAC consistently yields robust and substantial improvements across a diverse range of tasks and LLMs, offering compelling evidence of its efficacy. △ Less

Submitted 16 July, 2025; originally announced July 2025.

Comments: ACL 2025

KV-Latent: Dimensional-level KV Cache Reduction with Frequency-aware Rotary Positional Embedding

Authors: Luohe Shi, Zuchao Li, Lefei Zhang, Guoming Liu, Baoyuan Qi, Hai Zhao

Abstract: Large language models (LLMs) based on Transformer Decoders have become the preferred choice for conversational generative AI. Despite the overall superiority of the Decoder architecture, the gradually increasing Key-Value (KV) cache during inference has emerged as a primary efficiency bottleneck, both in aspects of memory consumption and data transfer bandwidth limitations. To address these challe… ▽ More Large language models (LLMs) based on Transformer Decoders have become the preferred choice for conversational generative AI. Despite the overall superiority of the Decoder architecture, the gradually increasing Key-Value (KV) cache during inference has emerged as a primary efficiency bottleneck, both in aspects of memory consumption and data transfer bandwidth limitations. To address these challenges, we propose a paradigm called KV-Latent. By down-sampling the Key-Value vector dimensions into a latent space, we can significantly reduce the KV Cache footprint and improve inference speed, only with a small amount of extra training, less than 1\% of pre-training takes. Besides, we enhanced the stability of Rotary Positional Embedding applied on lower-dimensional vectors by modifying its frequency sampling mechanism, avoiding noise introduced by higher frequencies while retaining position attenuation. Our experiments, including both models with Grouped Query Attention and those without, have yielded satisfactory results. Finally, we conducted comparative experiments to study the impact of separately reducing Key and Value components on model's performance. Our approach allows for the construction of more efficient language model systems, and opens the new possibility on KV Cache saving and efficient LLMs. Our code is available at https://github.com/ShiLuohe/KV-Latent. △ Less

Submitted 15 July, 2025; originally announced July 2025.

Comments: To be published in The 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025)

TOP: Trajectory Optimization via Parallel Optimization towards Constant Time Complexity

Authors: Jiajun Yu, Nanhe Chen, Guodong Liu, Chao Xu, Fei Gao, Yanjun Cao

Abstract: Optimization has been widely used to generate smooth trajectories for motion planning. However, existing trajectory optimization methods show weakness when dealing with large-scale long trajectories. Recent advances in parallel computing have accelerated optimization in some fields, but how to efficiently solve trajectory optimization via parallelism remains an open question. In this paper, we pro… ▽ More Optimization has been widely used to generate smooth trajectories for motion planning. However, existing trajectory optimization methods show weakness when dealing with large-scale long trajectories. Recent advances in parallel computing have accelerated optimization in some fields, but how to efficiently solve trajectory optimization via parallelism remains an open question. In this paper, we propose a novel trajectory optimization framework based on the Consensus Alternating Direction Method of Multipliers (CADMM) algorithm, which decomposes the trajectory into multiple segments and solves the subproblems in parallel. The proposed framework reduces the time complexity to O(1) per iteration to the number of segments, compared to O(N) of the state-of-the-art (SOTA) approaches. Furthermore, we introduce a closed-form solution that integrates convex linear and quadratic constraints to speed up the optimization, and we also present numerical solutions for general inequality constraints. A series of simulations and experiments demonstrate that our approach outperforms the SOTA approach in terms of efficiency and smoothness. Especially for a large-scale trajectory, with one hundred segments, achieving over a tenfold speedup. To fully explore the potential of our algorithm on modern parallel computing architectures, we deploy our framework on a GPU and show high performance with thousands of segments. △ Less

Submitted 16 July, 2025; v1 submitted 14 July, 2025; originally announced July 2025.

Comments: 8 pages, submitted to RA-L

TruckV2X: A Truck-Centered Perception Dataset

Authors: Tenghui Xie, Zhiying Song, Fuxi Wen, Jun Li, Guangzhao Liu, Zijian Zhao

Abstract: Autonomous trucking offers significant benefits, such as improved safety and reduced costs, but faces unique perception challenges due to trucks' large size and dynamic trailer movements. These challenges include extensive blind spots and occlusions that hinder the truck's perception and the capabilities of other road users. To address these limitations, cooperative perception emerges as a promisi… ▽ More Autonomous trucking offers significant benefits, such as improved safety and reduced costs, but faces unique perception challenges due to trucks' large size and dynamic trailer movements. These challenges include extensive blind spots and occlusions that hinder the truck's perception and the capabilities of other road users. To address these limitations, cooperative perception emerges as a promising solution. However, existing datasets predominantly feature light vehicle interactions or lack multi-agent configurations for heavy-duty vehicle scenarios. To bridge this gap, we introduce TruckV2X, the first large-scale truck-centered cooperative perception dataset featuring multi-modal sensing (LiDAR and cameras) and multi-agent cooperation (tractors, trailers, CAVs, and RSUs). We further investigate how trucks influence collaborative perception needs, establishing performance benchmarks while suggesting research priorities for heavy vehicle perception. The dataset provides a foundation for developing cooperative perception systems with enhanced occlusion handling capabilities, and accelerates the deployment of multi-agent autonomous trucking systems. The TruckV2X dataset is available at https://huggingface.co/datasets/XieTenghu1/TruckV2X. △ Less

Submitted 13 July, 2025; originally announced July 2025.

PLAME: Leveraging Pretrained Language Models to Generate Enhanced Protein Multiple Sequence Alignments

Authors: Hanqun Cao, Xinyi Zhou, Zijun Gao, Chenyu Wang, Xin Gao, Zhi Zhang, Chunbin Gu, Ge Liu, Pheng-Ann Heng

Abstract: Protein structure prediction is essential for drug discovery and understanding biological functions. While recent advancements like AlphaFold have achieved remarkable accuracy, most folding models rely heavily on multiple sequence alignments (MSAs) to boost prediction performance. This dependency limits their effectiveness on low-homology proteins and orphan proteins, where MSA information is spar… ▽ More Protein structure prediction is essential for drug discovery and understanding biological functions. While recent advancements like AlphaFold have achieved remarkable accuracy, most folding models rely heavily on multiple sequence alignments (MSAs) to boost prediction performance. This dependency limits their effectiveness on low-homology proteins and orphan proteins, where MSA information is sparse or unavailable. To address this limitation, we propose PLAME, a novel MSA design model that leverages evolutionary embeddings from pretrained protein language models. Unlike existing methods, PLAME introduces pretrained representations to enhance evolutionary information and employs a conservation-diversity loss to enhance generation quality. Additionally, we propose a novel MSA selection method to effectively screen high-quality MSAs and improve folding performance. We also propose a sequence quality assessment metric that provides an orthogonal perspective to evaluate MSA quality. On the AlphaFold2 benchmark of low-homology and orphan proteins, PLAME achieves state-of-the-art performance in folding enhancement and sequence quality assessment, with consistent improvements demonstrated on AlphaFold3. Ablation studies validate the effectiveness of the MSA selection method, while extensive case studies on various protein types provide insights into the relationship between AlphaFold's prediction quality and MSA characteristics. Furthermore, we demonstrate that PLAME can serve as an adapter achieving AlphaFold2-level accuracy with the ESMFold's inference speed. △ Less

Submitted 17 June, 2025; originally announced July 2025.

SpindleKV: A Novel KV Cache Reduction Method Balancing Both Shallow and Deep Layers

Authors: Zicong Tang, Shi Luohe, Zuchao Li, Baoyuan Qi, Guoming Liu, Lefei Zhang, Ping Wang

Abstract: Large Language Models (LLMs) have achieved impressive accomplishments in recent years. However, the increasing memory consumption of KV cache has possessed a significant challenge to the inference system. Eviction methods have revealed the inherent redundancy within the KV cache, demonstrating its potential for reduction, particularly in deeper layers. However, KV cache reduction for shallower lay… ▽ More Large Language Models (LLMs) have achieved impressive accomplishments in recent years. However, the increasing memory consumption of KV cache has possessed a significant challenge to the inference system. Eviction methods have revealed the inherent redundancy within the KV cache, demonstrating its potential for reduction, particularly in deeper layers. However, KV cache reduction for shallower layers has been found to be insufficient. Based on our observation that, the KV cache exhibits a high degree of similarity. Based on this observation, we proposed a novel KV cache reduction method, SpindleKV, which balances both shallow and deep layers. For deep layers, we employ an attention weight based eviction method, while for shallow layers, we apply a codebook based replacement approach which is learnt by similarity and merging policy. Moreover, SpindleKV addressed the Grouped-Query Attention (GQA) dilemma faced by other attention based eviction methods. Experiments on two common benchmarks with three different LLMs shown that SpindleKV obtained better KV cache reduction effect compared to baseline methods, while preserving similar or even better model performance. △ Less

Submitted 8 July, 2025; originally announced July 2025.

Comments: Accepted by ACL 2025 main

Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities

Authors: Gheorghe Comanici, Eric Bieber, Mike Schaekermann, Ice Pasupat, Noveen Sachdeva, Inderjit Dhillon, Marcel Blistein, Ori Ram, Dan Zhang, Evan Rosen, Luke Marris, Sam Petulla, Colin Gaffney, Asaf Aharoni, Nathan Lintz, Tiago Cardal Pais, Henrik Jacobsson, Idan Szpektor, Nan-Jiang Jiang, Krishna Haridasan, Ahmed Omran, Nikunj Saunshi, Dara Bahri, Gaurav Mishra, Eric Chu , et al. (3284 additional authors not shown)

Abstract: In this report, we introduce the Gemini 2.X model family: Gemini 2.5 Pro and Gemini 2.5 Flash, as well as our earlier Gemini 2.0 Flash and Flash-Lite models. Gemini 2.5 Pro is our most capable model yet, achieving SoTA performance on frontier coding and reasoning benchmarks. In addition to its incredible coding and reasoning skills, Gemini 2.5 Pro is a thinking model that excels at multimodal unde… ▽ More In this report, we introduce the Gemini 2.X model family: Gemini 2.5 Pro and Gemini 2.5 Flash, as well as our earlier Gemini 2.0 Flash and Flash-Lite models. Gemini 2.5 Pro is our most capable model yet, achieving SoTA performance on frontier coding and reasoning benchmarks. In addition to its incredible coding and reasoning skills, Gemini 2.5 Pro is a thinking model that excels at multimodal understanding and it is now able to process up to 3 hours of video content. Its unique combination of long context, multimodal and reasoning capabilities can be combined to unlock new agentic workflows. Gemini 2.5 Flash provides excellent reasoning abilities at a fraction of the compute and latency requirements and Gemini 2.0 Flash and Flash-Lite provide high performance at low latency and cost. Taken together, the Gemini 2.X model generation spans the full Pareto frontier of model capability vs cost, allowing users to explore the boundaries of what is possible with complex agentic problem solving. △ Less

Submitted 22 July, 2025; v1 submitted 7 July, 2025; originally announced July 2025.

Comments: 72 pages, 17 figures

Cross-Subject DD: A Cross-Subject Brain-Computer Interface Algorithm

Authors: Xiaoyuan Li, Xinru Xue, Bohan Zhang, Ye Sun, Shoushuo Xi, Gang Liu

Abstract: Brain-computer interface (BCI) based on motor imagery (MI) enables direct control of external devices by decoding the electroencephalogram (EEG) generated in the brain during imagined movements. However, due to inter-individual variability in brain activity, existing BCI models exhibit poor adaptability across subjects, thereby limiting their generalizability and widespread application. To address… ▽ More Brain-computer interface (BCI) based on motor imagery (MI) enables direct control of external devices by decoding the electroencephalogram (EEG) generated in the brain during imagined movements. However, due to inter-individual variability in brain activity, existing BCI models exhibit poor adaptability across subjects, thereby limiting their generalizability and widespread application. To address this issue, this paper proposes a cross-subject BCI algorithm named Cross-Subject DD (CSDD), which constructs a universal BCI model by extracting common features across subjects. The specific methods include: 1) training personalized models for each subject; 2) transforming personalized models into relation spectrums; 3) identifying common features through statistical analysis; and 4) constructing a cross-subject universal model based on common features. The experiments utilized the BCIC IV 2a dataset, involving nine subjects. Eight of these subjects were selected for training and extracing the common features, and the cross-subject decoding performance of the model was validated on the remaining subject. The results demonstrate that, compared with existing similar methods, our approach achieves a 3.28% improvement in performance. This paper introduces for the first time a novel method for extracting pure common features and constructing a universal cross-subject BCI model, thereby facilitating broader applications of BCI technology. △ Less

Submitted 2 July, 2025; originally announced July 2025.

Comments: 20 pages, 9 figures

IDAGC: Adaptive Generalized Human-Robot Collaboration via Human Intent Estimation and Multimodal Policy Learning

Authors: Haotian Liu, Yuchuang Tong, Guanchen Liu, Zhaojie Ju, Zhengtao Zhang

Abstract: In Human-Robot Collaboration (HRC), which encompasses physical interaction and remote cooperation, accurate estimation of human intentions and seamless switching of collaboration modes to adjust robot behavior remain paramount challenges. To address these issues, we propose an Intent-Driven Adaptive Generalized Collaboration (IDAGC) framework that leverages multimodal data and human intent estimat… ▽ More In Human-Robot Collaboration (HRC), which encompasses physical interaction and remote cooperation, accurate estimation of human intentions and seamless switching of collaboration modes to adjust robot behavior remain paramount challenges. To address these issues, we propose an Intent-Driven Adaptive Generalized Collaboration (IDAGC) framework that leverages multimodal data and human intent estimation to facilitate adaptive policy learning across multi-tasks in diverse scenarios, thereby facilitating autonomous inference of collaboration modes and dynamic adjustment of robotic actions. This framework overcomes the limitations of existing HRC methods, which are typically restricted to a single collaboration mode and lack the capacity to identify and transition between diverse states. Central to our framework is a predictive model that captures the interdependencies among vision, language, force, and robot state data to accurately recognize human intentions with a Conditional Variational Autoencoder (CVAE) and automatically switch collaboration modes. By employing dedicated encoders for each modality and integrating extracted features through a Transformer decoder, the framework efficiently learns multi-task policies, while force data optimizes compliance control and intent estimation accuracy during physical interactions. Experiments highlights our framework's practical potential to advance the comprehensive development of HRC. △ Less

Submitted 6 July, 2025; originally announced July 2025.

Comments: Accepted by IROS 2025

Hijacking JARVIS: Benchmarking Mobile GUI Agents against Unprivileged Third Parties

Authors: Guohong Liu, Jialei Ye, Jiacheng Liu, Yuanchun Li, Wei Liu, Pengzhi Gao, Jian Luan, Yunxin Liu

Abstract: Mobile GUI agents are designed to autonomously execute diverse device-control tasks by interpreting and interacting with mobile screens. Despite notable advancements, their resilience in real-world scenarios where screen content may be partially manipulated by untrustworthy third parties remains largely unexplored. Owing to their black-box and autonomous nature, these agents are vulnerable to mani… ▽ More Mobile GUI agents are designed to autonomously execute diverse device-control tasks by interpreting and interacting with mobile screens. Despite notable advancements, their resilience in real-world scenarios where screen content may be partially manipulated by untrustworthy third parties remains largely unexplored. Owing to their black-box and autonomous nature, these agents are vulnerable to manipulations that could compromise user devices. In this work, we present the first systematic investigation into the vulnerabilities of mobile GUI agents. We introduce a scalable attack simulation framework AgentHazard, which enables flexible and targeted modifications of screen content within existing applications. Leveraging this framework, we develop a comprehensive benchmark suite comprising both a dynamic task execution environment and a static dataset of vision-language-action tuples, totaling over 3,000 attack scenarios. The dynamic environment encompasses 58 reproducible tasks in an emulator with various types of hazardous UI content, while the static dataset is constructed from 210 screenshots collected from 14 popular commercial apps. Importantly, our content modifications are designed to be feasible for unprivileged third parties. We evaluate 7 widely-used mobile GUI agents and 5 common backbone models using our benchmark. Our findings reveal that all examined agents are significantly influenced by misleading third-party content (with an average misleading rate of 28.8% in human-crafted attack scenarios) and that their vulnerabilities are closely linked to the employed perception modalities and backbone LLMs. Furthermore, we assess training-based mitigation strategies, highlighting both the challenges and opportunities for enhancing the robustness of mobile GUI agents. Our code and data will be released at https://agenthazard.github.io. △ Less

Submitted 5 July, 2025; originally announced July 2025.

Test-Time Scaling with Reflective Generative Model

Authors: Zixiao Wang, Yuxin Wang, Xiaorui Wang, Mengting Xing, Jie Gao, Jianjun Xu, Guangcan Liu, Chenhui Jin, Zhuo Wang, Shengzhuo Zhang, Hongtao Xie

Abstract: We introduce our first reflective generative model MetaStone-S1, which obtains OpenAI o3-mini's performance via the new Reflective Generative Form. The new form focuses on high-quality reasoning trajectory selection and contains two novelties: 1) A unified interface for policy and process reward model: we share the backbone network and use task-specific heads for reasoning trajectory predicting an… ▽ More We introduce our first reflective generative model MetaStone-S1, which obtains OpenAI o3-mini's performance via the new Reflective Generative Form. The new form focuses on high-quality reasoning trajectory selection and contains two novelties: 1) A unified interface for policy and process reward model: we share the backbone network and use task-specific heads for reasoning trajectory predicting and scoring respectively, introducing only 53M extra parameters for trajectory scoring. 2) Eliminating the reliance on process-level annotation: we provide a self-supervised process reward model, which can directly learn the high-quality reasoning trajectory selection from the outcome reward. Equipped with the reflective generative form, MetaStone-S1 is naturally suitable for test-time scaling, and we provide three reasoning effort modes (low, medium, and high) based on the controllable thinking length. Experiments demonstrate that our MetaStone-S1 achieves comparable performance to OpenAI o3-mini's series with only 32B parameter size. To support the research community, we have open-sourced MetaStone-S1 at https://github.com/MetaStone-AI/MetaStone-S1. △ Less

Submitted 9 July, 2025; v1 submitted 2 July, 2025; originally announced July 2025.

IC-Custom: Diverse Image Customization via In-Context Learning

Authors: Yaowei Li, Xiaoyu Li, Zhaoyang Zhang, Yuxuan Bian, Gan Liu, Xinyuan Li, Jiale Xu, Wenbo Hu, Yating Liu, Lingen Li, Jing Cai, Yuexian Zou, Yancheng He, Ying Shan

Abstract: Image customization, a crucial technique for industrial media production, aims to generate content that is consistent with reference images. However, current approaches conventionally separate image customization into position-aware and position-free customization paradigms and lack a universal framework for diverse customization, limiting their applications across various scenarios. To overcome t… ▽ More Image customization, a crucial technique for industrial media production, aims to generate content that is consistent with reference images. However, current approaches conventionally separate image customization into position-aware and position-free customization paradigms and lack a universal framework for diverse customization, limiting their applications across various scenarios. To overcome these limitations, we propose IC-Custom, a unified framework that seamlessly integrates position-aware and position-free image customization through in-context learning. IC-Custom concatenates reference images with target images to a polyptych, leveraging DiT's multi-modal attention mechanism for fine-grained token-level interactions. We introduce the In-context Multi-Modal Attention (ICMA) mechanism with learnable task-oriented register tokens and boundary-aware positional embeddings to enable the model to correctly handle different task types and distinguish various inputs in polyptych configurations. To bridge the data gap, we carefully curated a high-quality dataset of 12k identity-consistent samples with 8k from real-world sources and 4k from high-quality synthetic data, avoiding the overly glossy and over-saturated synthetic appearance. IC-Custom supports various industrial applications, including try-on, accessory placement, furniture arrangement, and creative IP customization. Extensive evaluations on our proposed ProductBench and the publicly available DreamBench demonstrate that IC-Custom significantly outperforms community workflows, closed-source models, and state-of-the-art open-source approaches. IC-Custom achieves approximately 73% higher human preference across identity consistency, harmonicity, and text alignment metrics, while training only 0.4% of the original model parameters. Project page: https://liyaowei-stu.github.io/project/IC_Custom △ Less

Submitted 2 July, 2025; originally announced July 2025.

Comments: Project page: https://liyaowei-stu.github.io/project/IC_Custom

Discourse Heuristics For Paradoxically Moral Self-Correction

Authors: Guangliang Liu, Zimo Qi, Xitong Zhang, Kristen Marie Johnson

Abstract: Moral self-correction has emerged as a promising approach for aligning the output of Large Language Models (LLMs) with human moral values. However, moral self-correction techniques are subject to two primary paradoxes. First, despite empirical and theoretical evidence to support the effectiveness of self-correction, this LLM capability only operates at a superficial level. Second, while LLMs posse… ▽ More Moral self-correction has emerged as a promising approach for aligning the output of Large Language Models (LLMs) with human moral values. However, moral self-correction techniques are subject to two primary paradoxes. First, despite empirical and theoretical evidence to support the effectiveness of self-correction, this LLM capability only operates at a superficial level. Second, while LLMs possess the capability of self-diagnosing immoral aspects of their output, they struggle to identify the cause of this moral inconsistency during their self-correction process. To better understand and address these paradoxes, we analyze the discourse constructions in fine-tuning corpora designed to enhance moral self-correction, uncovering the existence of the heuristics underlying effective constructions. We demonstrate that moral self-correction relies on discourse constructions that reflect heuristic shortcuts, and that the presence of these heuristic shortcuts during self-correction leads to inconsistency when attempting to enhance both self-correction and self-diagnosis capabilities jointly. Based on our findings, we propose a solution to improve moral self-correction by leveraging the heuristics of curated datasets. We also highlight the generalization challenges of this capability, particularly in terms of learning from situated context and model scales. △ Less

Submitted 1 July, 2025; originally announced July 2025.

Transferable Modeling Strategies for Low-Resource LLM Tasks: A Prompt and Alignment-Based Approach

Authors: Shuangquan Lyu, Yingnan Deng, Guiran Liu, Zhen Qi, Ruotong Wang

Abstract: This paper addresses the limited transfer and adaptation capabilities of large language models in low-resource language scenarios. It proposes a unified framework that combines a knowledge transfer module with parameter-efficient fine-tuning strategies. The method introduces knowledge alignment loss and soft prompt tuning to guide the model in effectively absorbing the structural features of targe… ▽ More This paper addresses the limited transfer and adaptation capabilities of large language models in low-resource language scenarios. It proposes a unified framework that combines a knowledge transfer module with parameter-efficient fine-tuning strategies. The method introduces knowledge alignment loss and soft prompt tuning to guide the model in effectively absorbing the structural features of target languages or tasks under minimal annotation. This enhances both generalization performance and training stability. The framework includes lightweight adaptation modules to reduce computational costs. During training, it integrates freezing strategies and prompt injection to preserve the model's original knowledge while enabling quick adaptation to new tasks. The study also conducts stability analysis experiments and synthetic pseudo-data transfer experiments to systematically evaluate the method's applicability and robustness across different low-resource tasks. Experimental results show that compared with existing multilingual pre-trained models and mainstream transfer methods, the proposed approach achieves higher performance and stability on cross-lingual tasks such as MLQA, XQuAD, and PAWS-X. It demonstrates particularly strong advantages under extremely data-scarce conditions. The proposed method offers strong generality and scalability. It enhances task-specific adaptability while preserving the general capabilities of large language models. This makes it well-suited for complex semantic modeling and multilingual processing tasks. △ Less

Submitted 2 July, 2025; v1 submitted 1 July, 2025; originally announced July 2025.

PNAct: Crafting Backdoor Attacks in Safe Reinforcement Learning

Authors: Weiran Guo, Guanjun Liu, Ziyuan Zhou, Ling Wang

Abstract: Reinforcement Learning (RL) is widely used in tasks where agents interact with an environment to maximize rewards. Building on this foundation, Safe Reinforcement Learning (Safe RL) incorporates a cost metric alongside the reward metric, ensuring that agents adhere to safety constraints during decision-making. In this paper, we identify that Safe RL is vulnerable to backdoor attacks, which can man… ▽ More Reinforcement Learning (RL) is widely used in tasks where agents interact with an environment to maximize rewards. Building on this foundation, Safe Reinforcement Learning (Safe RL) incorporates a cost metric alongside the reward metric, ensuring that agents adhere to safety constraints during decision-making. In this paper, we identify that Safe RL is vulnerable to backdoor attacks, which can manipulate agents into performing unsafe actions. First, we introduce the relevant concepts and evaluation metrics for backdoor attacks in Safe RL. It is the first attack framework in the Safe RL field that involves both Positive and Negative Action sample (PNAct) is to implant backdoors, where positive action samples provide reference actions and negative action samples indicate actions to be avoided. We theoretically point out the properties of PNAct and design an attack algorithm. Finally, we conduct experiments to evaluate the effectiveness of our proposed backdoor attack framework, evaluating it with the established metrics. This paper highlights the potential risks associated with Safe RL and underscores the feasibility of such attacks. Our code and supplementary material are available at https://github.com/azure-123/PNAct. △ Less

Submitted 1 July, 2025; originally announced July 2025.

Region-Aware CAM: High-Resolution Weakly-Supervised Defect Segmentation via Salient Region Perception

Authors: Hang-Cheng Dong, Lu Zou, Bingguo Liu, Dong Ye, Guodong Liu

Abstract: Surface defect detection plays a critical role in industrial quality inspection. Recent advances in artificial intelligence have significantly enhanced the automation level of detection processes. However, conventional semantic segmentation and object detection models heavily rely on large-scale annotated datasets, which conflicts with the practical requirements of defect detection tasks. This pap… ▽ More Surface defect detection plays a critical role in industrial quality inspection. Recent advances in artificial intelligence have significantly enhanced the automation level of detection processes. However, conventional semantic segmentation and object detection models heavily rely on large-scale annotated datasets, which conflicts with the practical requirements of defect detection tasks. This paper proposes a novel weakly supervised semantic segmentation framework comprising two key components: a region-aware class activation map (CAM) and pseudo-label training. To address the limitations of existing CAM methods, especially low-resolution thermal maps, and insufficient detail preservation, we introduce filtering-guided backpropagation (FGBP), which refines target regions by filtering gradient magnitudes to identify areas with higher relevance to defects. Building upon this, we further develop a region-aware weighted module to enhance spatial precision. Finally, pseudo-label segmentation is implemented to refine the model's performance iteratively. Comprehensive experiments on industrial defect datasets demonstrate the superiority of our method. The proposed framework effectively bridges the gap between weakly supervised learning and high-precision defect segmentation, offering a practical solution for resource-constrained industrial scenarios. △ Less

Submitted 28 June, 2025; originally announced June 2025.

CaO$_2$: Rectifying Inconsistencies in Diffusion-Based Dataset Distillation

Authors: Haoxuan Wang, Zhenghao Zhao, Junyi Wu, Yuzhang Shang, Gaowen Liu, Yan Yan

Abstract: The recent introduction of diffusion models in dataset distillation has shown promising potential in creating compact surrogate datasets for large, high-resolution target datasets, offering improved efficiency and performance over traditional bi-level/uni-level optimization methods. However, current diffusion-based dataset distillation approaches overlook the evaluation process and exhibit two cri… ▽ More The recent introduction of diffusion models in dataset distillation has shown promising potential in creating compact surrogate datasets for large, high-resolution target datasets, offering improved efficiency and performance over traditional bi-level/uni-level optimization methods. However, current diffusion-based dataset distillation approaches overlook the evaluation process and exhibit two critical inconsistencies in the distillation process: (1) Objective Inconsistency, where the distillation process diverges from the evaluation objective, and (2) Condition Inconsistency, leading to mismatches between generated images and their corresponding conditions. To resolve these issues, we introduce Condition-aware Optimization with Objective-guided Sampling (CaO$_2$), a two-stage diffusion-based framework that aligns the distillation process with the evaluation objective. The first stage employs a probability-informed sample selection pipeline, while the second stage refines the corresponding latent representations to improve conditional likelihood. CaO$_2$ achieves state-of-the-art performance on ImageNet and its subsets, surpassing the best-performing baselines by an average of 2.3% accuracy. △ Less

Submitted 8 July, 2025; v1 submitted 27 June, 2025; originally announced June 2025.

Comments: ICCV 2025. Code is available at https://github.com/hatchetProject/CaO2

Adjoint Schrödinger Bridge Sampler

Authors: Guan-Horng Liu, Jaemoo Choi, Yongxin Chen, Benjamin Kurt Miller, Ricky T. Q. Chen

Abstract: Computational methods for learning to sample from the Boltzmann distribution -- where the target distribution is known only up to an unnormalized energy function -- have advanced significantly recently. Due to the lack of explicit target samples, however, prior diffusion-based methods, known as diffusion samplers, often require importance-weighted estimation or complicated learning processes. Both… ▽ More Computational methods for learning to sample from the Boltzmann distribution -- where the target distribution is known only up to an unnormalized energy function -- have advanced significantly recently. Due to the lack of explicit target samples, however, prior diffusion-based methods, known as diffusion samplers, often require importance-weighted estimation or complicated learning processes. Both trade off scalability with extensive evaluations of the energy and model, thereby limiting their practical usage. In this work, we propose Adjoint Schrödinger Bridge Sampler (ASBS), a new diffusion sampler that employs simple and scalable matching-based objectives yet without the need to estimate target samples during training. ASBS is grounded on a mathematical model -- the Schrödinger Bridge -- which enhances sampling efficiency via kinetic-optimal transportation. Through a new lens of stochastic optimal control theory, we demonstrate how SB-based diffusion samplers can be learned at scale via Adjoint Matching and prove convergence to the global solution. Notably, ASBS generalizes the recent Adjoint Sampling (Havens et al., 2025) to arbitrary source distributions by relaxing the so-called memoryless condition that largely restricts the design space. Through extensive experiments, we demonstrate the effectiveness of ASBS on sampling from classical energy functions, amortized conformer generation, and molecular Boltzmann distributions. △ Less

Submitted 27 June, 2025; originally announced June 2025.

Non-equilibrium Annealed Adjoint Sampler

Authors: Jaemoo Choi, Yongxin Chen, Molei Tao, Guan-Horng Liu

Abstract: Recently, there has been significant progress in learning-based diffusion samplers, which aim to sample from a given unnormalized density. These methods typically follow one of two paradigms: (i) formulating sampling as an unbiased stochastic optimal control (SOC) problem using a canonical reference process, or (ii) refining annealed path measures through importance-weighted sampling. Although ann… ▽ More Recently, there has been significant progress in learning-based diffusion samplers, which aim to sample from a given unnormalized density. These methods typically follow one of two paradigms: (i) formulating sampling as an unbiased stochastic optimal control (SOC) problem using a canonical reference process, or (ii) refining annealed path measures through importance-weighted sampling. Although annealing approaches have advantages in guiding samples toward high-density regions, reliance on importance sampling leads to high variance and limited scalability in practice. In this paper, we introduce the \textbf{Non-equilibrium Annealed Adjoint Sampler (NAAS)}, a novel SOC-based diffusion sampler that leverages annealed reference dynamics without resorting to importance sampling. NAAS employs a lean adjoint system inspired by adjoint matching, enabling efficient and scalable training. We demonstrate the effectiveness of our approach across a range of tasks, including sampling from classical energy landscapes and molecular Boltzmann distribution. △ Less

Submitted 25 June, 2025; v1 submitted 22 June, 2025; originally announced June 2025.

Comments: 21 pages, 7 figures

RoboTwin 2.0: A Scalable Data Generator and Benchmark with Strong Domain Randomization for Robust Bimanual Robotic Manipulation

Authors: Tianxing Chen, Zanxin Chen, Baijun Chen, Zijian Cai, Yibin Liu, Qiwei Liang, Zixuan Li, Xianliang Lin, Yiheng Ge, Zhenyu Gu, Weiliang Deng, Yubin Guo, Tian Nian, Xuanbing Xie, Qiangyu Chen, Kailun Su, Tianling Xu, Guodong Liu, Mengkang Hu, Huan-ang Gao, Kaixuan Wang, Zhixuan Liang, Yusen Qin, Xiaokang Yang, Ping Luo , et al. (1 additional authors not shown)

Abstract: Simulation-based data synthesis has emerged as a powerful paradigm for enhancing real-world robotic manipulation. However, existing synthetic datasets remain insufficient for robust bimanual manipulation due to two challenges: (1) the lack of an efficient, scalable data generation method for novel tasks, and (2) oversimplified simulation environments that fail to capture real-world complexity. We… ▽ More Simulation-based data synthesis has emerged as a powerful paradigm for enhancing real-world robotic manipulation. However, existing synthetic datasets remain insufficient for robust bimanual manipulation due to two challenges: (1) the lack of an efficient, scalable data generation method for novel tasks, and (2) oversimplified simulation environments that fail to capture real-world complexity. We present RoboTwin 2.0, a scalable simulation framework that enables automated, large-scale generation of diverse and realistic data, along with unified evaluation protocols for dual-arm manipulation. We first construct RoboTwin-OD, a large-scale object library comprising 731 instances across 147 categories, each annotated with semantic and manipulation-relevant labels. Building on this foundation, we develop an expert data synthesis pipeline that combines multimodal large language models (MLLMs) with simulation-in-the-loop refinement to generate task-level execution code automatically. To improve sim-to-real transfer, RoboTwin 2.0 incorporates structured domain randomization along five axes: clutter, lighting, background, tabletop height and language instructions, thereby enhancing data diversity and policy robustness. We instantiate this framework across 50 dual-arm tasks spanning five robot embodiments, and pre-collect over 100,000 domain-randomized expert trajectories. Empirical results show a 10.9% gain in code generation success and improved generalization to novel real-world scenarios. A VLA model fine-tuned on our dataset achieves a 367% relative improvement (42.0% vs. 9.0%) on unseen scene real-world tasks, while zero-shot models trained solely on our synthetic data achieve a 228% relative gain, highlighting strong generalization without real-world supervision. We release the data generator, benchmark, dataset, and code to support scalable research in robust bimanual manipulation. △ Less

Submitted 22 June, 2025; originally announced June 2025.

Comments: Project Page: https://robotwin-platform.github.io/

Minimum-Weight Half-Plane Hitting Set

Authors: Gang Liu, Haitao Wang

Abstract: Given a set $P$ of $n$ weighted points and a set $H$ of $n$ half-planes in the plane, the hitting set problem is to compute a subset $P'$ of points from $P$ such that each half-plane contains at least one point from $P'$ and the total weight of the points in $P'$ is minimized. The previous best algorithm solves the problem in $O(n^{7/2}\log^2 n)$ time. In this paper, we present a new algorithm wit… ▽ More Given a set $P$ of $n$ weighted points and a set $H$ of $n$ half-planes in the plane, the hitting set problem is to compute a subset $P'$ of points from $P$ such that each half-plane contains at least one point from $P'$ and the total weight of the points in $P'$ is minimized. The previous best algorithm solves the problem in $O(n^{7/2}\log^2 n)$ time. In this paper, we present a new algorithm with runtime $O(n^{5/2}\log^2 n)$. △ Less

Submitted 20 June, 2025; originally announced June 2025.

Comments: To appear in CCCG 2025. arXiv admin note: text overlap with arXiv:2407.00329, arXiv:2501.02195

Autonomous Trajectory Optimization for UAVs in Disaster Zone Using Henry Gas Optimization Scheme

Authors: Zakria Qadir, Muhammad Bilal, Guoqiang Liu, Xiaolong Xu

Abstract: The unmanned aerial vehicles (UAVs) in a disaster-prone environment plays important role in assisting the rescue services and providing the internet connectivity with the outside world. However, in such a complex environment the selection of optimum trajectory of UAVs is of utmost importance. UAV trajectory optimization deals with finding the shortest path in the minimal possible time. In this pap… ▽ More The unmanned aerial vehicles (UAVs) in a disaster-prone environment plays important role in assisting the rescue services and providing the internet connectivity with the outside world. However, in such a complex environment the selection of optimum trajectory of UAVs is of utmost importance. UAV trajectory optimization deals with finding the shortest path in the minimal possible time. In this paper, a cluster optimization scheme (COS) is proposed using the Henry gas optimization (HGO) metaheuristic algorithm to identify the shortest path having minimal transportation cost and algorithm complexity. The mathematical model is designed for COS using the HGO algorithm and compared with the state-of-the-art metaheuristic algorithms such as particle swarm optimization (PSO), grey wolf optimization (GWO), cuckoo search algorithm (CSA) and barnacles mating optimizer (BMO). In order to prove the robustness of the proposed model, four different scenarios are evaluated that includes ambient environment, constrict environment, tangled environment, and complex environment. In all the aforementioned scenarios, the HGO algorithm outperforms the existing algorithms. Particularly, in the ambient environment, the HGO algorithm achieves a 39.3% reduction in transportation cost and a 16.8% reduction in computational time as compared to the PSO algorithm. Hence, the HGO algorithm can be used for autonomous trajectory optimization of UAVs in smart cities. △ Less

Submitted 18 June, 2025; originally announced June 2025.

Comments: 12 pages, 9 figuers

ACM Class: C.2; I.6

Unsupervised deep learning model for fast energy layer pre-selection of delivery-efficient proton arc therapy plan optimization of nasopharyngeal carcinoma

Authors: Bohan Yang, Gang Liu, Rirao Dao, Yujia Qian, Ke Shi, Anke Tang, Yong Luo, Jingnan Liu

Abstract: Objective. Proton arc therapy (PAT) is an emerging and promising modality in radiotherapy, offering several advantages over conventional intensitymodulated proton therapy (IMPT). However, identifying the optimal energy layer (EL) sequence remains computationally intensive due to the large number of possible energy layer transitions. This study proposes an unsupervised deep learning framework for f… ▽ More Objective. Proton arc therapy (PAT) is an emerging and promising modality in radiotherapy, offering several advantages over conventional intensitymodulated proton therapy (IMPT). However, identifying the optimal energy layer (EL) sequence remains computationally intensive due to the large number of possible energy layer transitions. This study proposes an unsupervised deep learning framework for fast and effective EL pre-selection, aiming to minimize energy layer switch time while preserving high plan quality. Approach. We introduce a novel data representation method, spot-count representation, which encodes the number of proton spots intersecting the target and organs at risk (OARs) in a matrix structured by sorted gantry angles and energy layers. This representation is the input of a UNet-based architecture, SPArcdl, which is trained to optimize a tri-objective function: maximizing target coverage, minimizing OAR exposure, and reducing energy switching time. The model is evaluated on 54 nasopharyngeal cancer cases, and its performance is benchmarked against plans generated by SPArcparticle swarm. Main results. SPArcdl produces EL pre-selection that significantly improves both plan quality and delivery efficiency. Compared to SPArc particle swarm, it enhances the conformity index by 0.16 (p < 0.01), reduces the homogeneity index by 0.71 (p < 0.01), shortens the energy switching time by 38.4% (p < 0.01), and lowers the mean dose to brainstem by 0.21 (p < 0.01). The results unintentionally reveal employing unchanged ELS is more time-wise efficient than descended ELS. SPArcdl's inference time is within 1 second. Significance. SPArcdl is a fast and effective tool for generating high-quality PAT plans by strategically pre-selecting energy layers to reduce delivery time while maintaining excellent dosimetric performance. △ Less

Submitted 18 June, 2025; originally announced June 2025.

Smartphone-integrated RPA-CRISPR-Cas12a Detection System with Microneedle Sampling for Point-of-Care Diagnosis of Potato Late Blight in Early Stage

Authors: Jiangnan Zhao, Hanbo Xu, Cifu Xu, Wenlong Yin, Laixin Luo, Gang Liu, Yan Wang

Abstract: Potato late blight, caused by the oomycete pathogen Phytophthora infestans, is one of the most devastating diseases affecting potato crops in the history. Although conventional detection methods of plant diseases such as PCR and LAMP are highly sensitive and specific, they rely on bulky and expensive laboratory equipment and involve complex operations, making them impracticable for point-of care d… ▽ More Potato late blight, caused by the oomycete pathogen Phytophthora infestans, is one of the most devastating diseases affecting potato crops in the history. Although conventional detection methods of plant diseases such as PCR and LAMP are highly sensitive and specific, they rely on bulky and expensive laboratory equipment and involve complex operations, making them impracticable for point-of care diagnosis in the field. Here in this study, we report a portable RPA-CRISPR based diagnosis system for plant disease, integrating smartphone for acquisition and analysis of fluorescent images. A polyvinyl alcohol (PVA) microneedle patch was employed for sample extraction on the plant leaves within one minute, the DNA extraction efficiency achieved 56 ug/mg, which is approximately 3 times to the traditional CTAB methods (18 ug/mg). The system of RPA-CRISPR-Cas12a isothermal assay was established to specifically target P. infestans with no cross-reactivity observed against closely-related species (P. sojae, P. capsici). The system demonstrated a detection limit of 2 pg/uL for P. infestans genomic DNA, offering sensitivity comparable to that of benchtop laboratory equipment. The system demonstrates the early-stage diagnosis capability by achieving a approximately 80% and 100% detection rate on the third and fourth day post-inoculation respectively, before visible symptoms observed on the leaves. The smartphone-based "sample-to-result" system decouples the limitations of traditional methods that rely heavily on specialized equipment, offering a promising way for early-stage plant disease detection and control in the field. △ Less

Submitted 10 June, 2025; originally announced June 2025.

Comments: 32 pages,7 figures,1 table

Truncated Proximal Policy Optimization

Authors: Tiantian Fan, Lingjun Liu, Yu Yue, Jiaze Chen, Chengyi Wang, Qiying Yu, Chi Zhang, Zhiqi Lin, Ruofei Zhu, Yufeng Yuan, Xiaochen Zuo, Bole Ma, Mofan Zhang, Gaohong Liu, Ru Zhang, Haotian Zhou, Cong Xie, Ruidong Zhu, Zhi Zhang, Xin Liu, Mingxuan Wang, Lin Yan, Yonghui Wu

Abstract: Recently, test-time scaling Large Language Models (LLMs) have demonstrated exceptional reasoning capabilities across scientific and professional tasks by generating long chains-of-thought (CoT). As a crucial component for developing these reasoning models, reinforcement learning (RL), exemplified by Proximal Policy Optimization (PPO) and its variants, allows models to learn through trial and error… ▽ More Recently, test-time scaling Large Language Models (LLMs) have demonstrated exceptional reasoning capabilities across scientific and professional tasks by generating long chains-of-thought (CoT). As a crucial component for developing these reasoning models, reinforcement learning (RL), exemplified by Proximal Policy Optimization (PPO) and its variants, allows models to learn through trial and error. However, PPO can be time-consuming due to its inherent on-policy nature, which is further exacerbated by increasing response lengths. In this work, we propose Truncated Proximal Policy Optimization (T-PPO), a novel extension to PPO that improves training efficiency by streamlining policy update and length-restricted response generation. T-PPO mitigates the issue of low hardware utilization, an inherent drawback of fully synchronized long-generation procedures, where resources often sit idle during the waiting periods for complete rollouts. Our contributions are two-folds. First, we propose Extended Generalized Advantage Estimation (EGAE) for advantage estimation derived from incomplete responses while maintaining the integrity of policy learning. Second, we devise a computationally optimized mechanism that allows for the independent optimization of the policy and value models. By selectively filtering prompt and truncated tokens, this mechanism reduces redundant computations and accelerates the training process without sacrificing convergence performance. We demonstrate the effectiveness and efficacy of T-PPO on AIME 2024 with a 32B base model. The experimental results show that T-PPO improves the training efficiency of reasoning LLMs by up to 2.5x and outperforms its existing competitors. △ Less

Submitted 17 June, 2025; originally announced June 2025.

Exploring Diffusion with Test-Time Training on Efficient Image Restoration

Authors: Rongchang Lu, Tianduo Luo, Yunzhi Jiang, Conghan Yue, Pei Yang, Guibao Liu, Changyang Gu

Abstract: Image restoration faces challenges including ineffective feature fusion, computational bottlenecks and inefficient diffusion processes. To address these, we propose DiffRWKVIR, a novel framework unifying Test-Time Training (TTT) with efficient diffusion. Our approach introduces three key innovations: (1) Omni-Scale 2D State Evolution extends RWKV's location-dependent parameterization to hierarchic… ▽ More Image restoration faces challenges including ineffective feature fusion, computational bottlenecks and inefficient diffusion processes. To address these, we propose DiffRWKVIR, a novel framework unifying Test-Time Training (TTT) with efficient diffusion. Our approach introduces three key innovations: (1) Omni-Scale 2D State Evolution extends RWKV's location-dependent parameterization to hierarchical multi-directional 2D scanning, enabling global contextual awareness with linear complexity O(L); (2) Chunk-Optimized Flash Processing accelerates intra-chunk parallelism by 3.2x via contiguous chunk processing (O(LCd) complexity), reducing sequential dependencies and computational overhead; (3) Prior-Guided Efficient Diffusion extracts a compact Image Prior Representation (IPR) in only 5-20 steps, proving 45% faster training/inference than DiffIR while solving computational inefficiency in denoising. Evaluated across super-resolution and inpainting benchmarks (Set5, Set14, BSD100, Urban100, Places365), DiffRWKVIR outperforms SwinIR, HAT, and MambaIR/v2 in PSNR, SSIM, LPIPS, and efficiency metrics. Our method establishes a new paradigm for adaptive, high-efficiency image restoration with optimized hardware utilization. △ Less

Submitted 22 June, 2025; v1 submitted 17 June, 2025; originally announced June 2025.

ACM Class: I.4.9

Infinity Instruct: Scaling Instruction Selection and Synthesis to Enhance Language Models

Authors: Jijie Li, Li Du, Hanyu Zhao, Bo-wen Zhang, Liangdong Wang, Boyan Gao, Guang Liu, Yonghua Lin

Abstract: Large Language Models (LLMs) demonstrate strong performance in real-world applications, yet existing open-source instruction datasets often concentrate on narrow domains, such as mathematics or coding, limiting generalization and widening the gap with proprietary models. To bridge this gap, we introduce Infinity-Instruct, a high-quality instruction dataset designed to enhance both foundational and… ▽ More Large Language Models (LLMs) demonstrate strong performance in real-world applications, yet existing open-source instruction datasets often concentrate on narrow domains, such as mathematics or coding, limiting generalization and widening the gap with proprietary models. To bridge this gap, we introduce Infinity-Instruct, a high-quality instruction dataset designed to enhance both foundational and chat capabilities of LLMs through a two-phase pipeline. In Phase 1, we curate 7.4M high-quality foundational instructions (InfInstruct-F-7.4M) from over 100M samples using hybrid data selection techniques. In Phase 2, we synthesize 1.5M high-quality chat instructions (InfInstruct-G-1.5M) through a two-stage process involving instruction selection, evolution, and diagnostic filtering. We empirically evaluate Infinity-Instruct by fine-tuning several open-source models, including Mistral, LLaMA, Qwen, and Yi, and observe substantial performance gains across both foundational and instruction following benchmarks, consistently surpassing official instruction-tuned counterparts. Notably, InfInstruct-LLaMA3.1-70B outperforms GPT-4-0314 by 8.6\% on instruction following tasks while achieving comparable foundational performance. These results underscore the synergy between foundational and chat training and offer new insights into holistic LLM development. Our dataset\footnote{https://huggingface.co/datasets/BAAI/Infinity-Instruct} and codes\footnote{https://gitee.com/li-touch/infinity-instruct} have been publicly released. △ Less

Submitted 9 June, 2025; originally announced June 2025.

AIR: Zero-shot Generative Model Adaptation with Iterative Refinement

Authors: Guimeng Liu, Milad Abdollahzadeh, Ngai-Man Cheung

Abstract: Zero-shot generative model adaptation (ZSGM) aims to adapt a pre-trained generator to a target domain using only text guidance and without any samples from the target domain. Central to recent ZSGM approaches are directional loss which use the text guidance in the form of aligning the image offset with text offset in the embedding space of a vision-language model like CLIP. This is similar to the… ▽ More Zero-shot generative model adaptation (ZSGM) aims to adapt a pre-trained generator to a target domain using only text guidance and without any samples from the target domain. Central to recent ZSGM approaches are directional loss which use the text guidance in the form of aligning the image offset with text offset in the embedding space of a vision-language model like CLIP. This is similar to the analogical reasoning in NLP where the offset between one pair of words is used to identify a missing element in another pair by aligning the offset between these two pairs. However, a major limitation of existing ZSGM methods is that the learning objective assumes the complete alignment between image offset and text offset in the CLIP embedding space, resulting in quality degrade in generated images. Our work makes two main contributions. Inspired by the offset misalignment studies in NLP, as our first contribution, we perform an empirical study to analyze the misalignment between text offset and image offset in CLIP embedding space for various large publicly available datasets. Our important finding is that offset misalignment in CLIP embedding space is correlated with concept distance, i.e., close concepts have a less offset misalignment. To address the limitations of the current approaches, as our second contribution, we propose Adaptation with Iterative Refinement (AIR) which is the first ZSGM approach to focus on improving target domain image quality based on our new insight on offset misalignment.Qualitative, quantitative, and user study in 26 experiment setups consistently demonstrate the proposed AIR approach achieves SOTA performance. Additional experiments are in Supp. △ Less

Submitted 12 June, 2025; originally announced June 2025.

Momentum Multi-Marginal Schrödinger Bridge Matching

Authors: Panagiotis Theodoropoulos, Augustinos D. Saravanos, Evangelos A. Theodorou, Guan-Horng Liu

Abstract: Understanding complex systems by inferring trajectories from sparse sample snapshots is a fundamental challenge in a wide range of domains, e.g., single-cell biology, meteorology, and economics. Despite advancements in Bridge and Flow matching frameworks, current methodologies rely on pairwise interpolation between adjacent snapshots. This hinders their ability to capture long-range temporal depen… ▽ More Understanding complex systems by inferring trajectories from sparse sample snapshots is a fundamental challenge in a wide range of domains, e.g., single-cell biology, meteorology, and economics. Despite advancements in Bridge and Flow matching frameworks, current methodologies rely on pairwise interpolation between adjacent snapshots. This hinders their ability to capture long-range temporal dependencies and potentially affects the coherence of the inferred trajectories. To address these issues, we introduce \textbf{Momentum Multi-Marginal Schrödinger Bridge Matching (3MSBM)}, a novel matching framework that learns smooth measure-valued splines for stochastic systems that satisfy multiple positional constraints. This is achieved by lifting the dynamics to phase space and generalizing stochastic bridges to be conditioned on several points, forming a multi-marginal conditional stochastic optimal control problem. The underlying dynamics are then learned by minimizing a variational objective, having fixed the path induced by the multi-marginal conditional bridge. As a matching approach, 3MSBM learns transport maps that preserve intermediate marginals throughout training, significantly improving convergence and scalability. Extensive experimentation in a series of real-world applications validates the superior performance of 3MSBM compared to existing methods in capturing complex dynamics with temporal dependencies, opening new avenues for training matching frameworks in multi-marginal settings. △ Less

Submitted 11 June, 2025; originally announced June 2025.

STREAMINGGS: Voxel-Based Streaming 3D Gaussian Splatting with Memory Optimization and Architectural Support

Authors: Chenqi Zhang, Yu Feng, Jieru Zhao, Guangda Liu, Wenchao Ding, Chentao Wu, Minyi Guo

Abstract: 3D Gaussian Splatting (3DGS) has gained popularity for its efficiency and sparse Gaussian-based representation. However, 3DGS struggles to meet the real-time requirement of 90 frames per second (FPS) on resource-constrained mobile devices, achieving only 2 to 9 FPS.Existing accelerators focus on compute efficiency but overlook memory efficiency, leading to redundant DRAM traffic. We introduce STRE… ▽ More 3D Gaussian Splatting (3DGS) has gained popularity for its efficiency and sparse Gaussian-based representation. However, 3DGS struggles to meet the real-time requirement of 90 frames per second (FPS) on resource-constrained mobile devices, achieving only 2 to 9 FPS.Existing accelerators focus on compute efficiency but overlook memory efficiency, leading to redundant DRAM traffic. We introduce STREAMINGGS, a fully streaming 3DGS algorithm-architecture co-design that achieves fine-grained pipelining and reduces DRAM traffic by transforming from a tile-centric rendering to a memory-centric rendering. Results show that our design achieves up to 45.7 $\times$ speedup and 62.9 $\times$ energy savings over mobile Ampere GPUs. △ Less

Submitted 9 June, 2025; originally announced June 2025.

Language-Vision Planner and Executor for Text-to-Visual Reasoning

Authors: Yichang Xu, Gaowen Liu, Ramana Rao Kompella, Sihao Hu, Tiansheng Huang, Fatih Ilhan, Selim Furkan Tekin, Zachary Yahn, Ling Liu

Abstract: The advancement in large language models (LLMs) and large vision models has fueled the rapid progress in multi-modal visual-text reasoning capabilities. However, existing vision-language models (VLMs) to date suffer from generalization performance. Inspired by recent development in LLMs for visual reasoning, this paper presents VLAgent, an AI system that can create a step-by-step visual reasoning… ▽ More The advancement in large language models (LLMs) and large vision models has fueled the rapid progress in multi-modal visual-text reasoning capabilities. However, existing vision-language models (VLMs) to date suffer from generalization performance. Inspired by recent development in LLMs for visual reasoning, this paper presents VLAgent, an AI system that can create a step-by-step visual reasoning plan with an easy-to-understand script and execute each step of the plan in real time by integrating planning script with execution verifications via an automated process supported by VLAgent. In the task planning phase, VLAgent fine-tunes an LLM through in-context learning to generate a step-by-step planner for each user-submitted text-visual reasoning task. During the plan execution phase, VLAgent progressively refines the composition of neuro-symbolic executable modules to generate high-confidence reasoning results. VLAgent has three unique design characteristics: First, we improve the quality of plan generation through in-context learning, improving logic reasoning by reducing erroneous logic steps, incorrect programs, and LLM hallucinations. Second, we design a syntax-semantics parser to identify and correct additional logic errors of the LLM-generated planning script prior to launching the plan executor. Finally, we employ the ensemble method to improve the generalization performance of our step-executor. Extensive experiments with four visual reasoning benchmarks (GQA, MME, NLVR2, VQAv2) show that VLAgent achieves significant performance enhancement for multimodal text-visual reasoning applications, compared to the exiting representative VLMs and LLM based visual composition approaches like ViperGPT and VisProg, thanks to the novel optimization modules of VLAgent back-engine (SS-Parser, Plan Repairer, Output Verifiers). Code and data will be made available upon paper acceptance. △ Less

Submitted 9 June, 2025; originally announced June 2025.

Fast ECoT: Efficient Embodied Chain-of-Thought via Thoughts Reuse

Authors: Zhekai Duan, Yuan Zhang, Shikai Geng, Gaowen Liu, Joschka Boedecker, Chris Xiaoxuan Lu

Abstract: Embodied Chain-of-Thought (ECoT) reasoning enhances vision-language-action (VLA) models by improving performance and interpretability through intermediate reasoning steps. However, its sequential autoregressive token generation introduces significant inference latency, limiting real-time deployment. We propose Fast ECoT, an inference-time acceleration method that exploits the structured and repeti… ▽ More Embodied Chain-of-Thought (ECoT) reasoning enhances vision-language-action (VLA) models by improving performance and interpretability through intermediate reasoning steps. However, its sequential autoregressive token generation introduces significant inference latency, limiting real-time deployment. We propose Fast ECoT, an inference-time acceleration method that exploits the structured and repetitive nature of ECoT to (1) cache and reuse high-level reasoning across timesteps and (2) parallelise the generation of modular reasoning steps. Additionally, we introduce an asynchronous scheduler that decouples reasoning from action decoding, further boosting responsiveness. Fast ECoT requires no model changes or additional training and integrates easily into existing VLA pipelines. Experiments in both simulation (LIBERO) and real-world robot tasks show up to a 7.5% reduction in latency with comparable or improved task success rate and reasoning faithfulness, bringing ECoT policies closer to practical real-time deployment. △ Less

Submitted 9 June, 2025; originally announced June 2025.

Curriculum Learning With Counterfactual Group Relative Policy Advantage For Multi-Agent Reinforcement Learning

Authors: Weiqiang Jin, Hongyang Du, Guizhong Liu, Dong In Kim

Abstract: Multi-agent reinforcement learning (MARL) has achieved strong performance in cooperative adversarial tasks. However, most existing methods typically train agents against fixed opponent strategies and rely on such meta-static difficulty conditions, which limits their adaptability to changing environments and often leads to suboptimal policies. Inspired by the success of curriculum learning (CL) in… ▽ More Multi-agent reinforcement learning (MARL) has achieved strong performance in cooperative adversarial tasks. However, most existing methods typically train agents against fixed opponent strategies and rely on such meta-static difficulty conditions, which limits their adaptability to changing environments and often leads to suboptimal policies. Inspired by the success of curriculum learning (CL) in supervised tasks, we propose a dynamic CL framework for MARL that employs an self-adaptive difficulty adjustment mechanism. This mechanism continuously modulates opponent strength based on real-time agent training performance, allowing agents to progressively learn from easier to more challenging scenarios. However, the dynamic nature of CL introduces instability due to nonstationary environments and sparse global rewards. To address this challenge, we develop a Counterfactual Group Relative Policy Advantage (CGRPA), which is tightly coupled with the curriculum by providing intrinsic credit signals that reflect each agent's impact under evolving task demands. CGRPA constructs a counterfactual advantage function that isolates individual contributions within group behavior, facilitating more reliable policy updates throughout the curriculum. CGRPA evaluates each agent's contribution through constructing counterfactual action advantage function, providing intrinsic rewards that enhance credit assignment and stabilize learning under non-stationary conditions. Extensive experiments demonstrate that our method improves both training stability and final performance, achieving competitive results against state-of-the-art methods. The code is available at https://github.com/NICE-HKU/CL2MARL-SMAC. △ Less

Submitted 9 June, 2025; originally announced June 2025.

Comments: 16 pages; 12figures

CCI4.0: A Bilingual Pretraining Dataset for Enhancing Reasoning in Large Language Models

Authors: Guang Liu, Liangdong Wang, Jijie Li, Yang Yu, Yao Xu, Jiabei Chen, Yu Bai, Feng Liao, Yonghua Lin

Abstract: We introduce CCI4.0, a large-scale bilingual pre-training dataset engineered for superior data quality and diverse human-like reasoning trajectory. CCI4.0 occupies roughly $35$ TB of disk space and comprises two sub-datasets: CCI4.0-M2-Base and CCI4.0-M2-CoT. CCI4.0-M2-Base combines a $5.2$ TB carefully curated Chinese web corpus, a $22.5$ TB English subset from Nemotron-CC, and diverse sources fr… ▽ More We introduce CCI4.0, a large-scale bilingual pre-training dataset engineered for superior data quality and diverse human-like reasoning trajectory. CCI4.0 occupies roughly $35$ TB of disk space and comprises two sub-datasets: CCI4.0-M2-Base and CCI4.0-M2-CoT. CCI4.0-M2-Base combines a $5.2$ TB carefully curated Chinese web corpus, a $22.5$ TB English subset from Nemotron-CC, and diverse sources from math, wiki, arxiv, and code. Although these data are mostly sourced from well-processed datasets, the quality standards of various domains are dynamic and require extensive expert experience and labor to process. So, we propose a novel pipeline justifying data quality mainly based on models through two-stage deduplication, multiclassifier quality scoring, and domain-aware fluency filtering. We extract $4.5$ billion pieces of CoT(Chain-of-Thought) templates, named CCI4.0-M2-CoT. Differing from the distillation of CoT from larger models, our proposed staged CoT extraction exemplifies diverse reasoning patterns and significantly decreases the possibility of hallucination. Empirical evaluations demonstrate that LLMs pre-trained in CCI4.0 benefit from cleaner, more reliable training signals, yielding consistent improvements in downstream tasks, especially in math and code reflection tasks. Our results underscore the critical role of rigorous data curation and human thinking templates in advancing LLM performance, shedding some light on automatically processing pretraining corpora. △ Less

Submitted 9 June, 2025; originally announced June 2025.

ProteinZero: Self-Improving Protein Generation via Online Reinforcement Learning

Authors: Ziwen Wang, Jiajun Fan, Ruihan Guo, Thao Nguyen, Heng Ji, Ge Liu

Abstract: Protein generative models have shown remarkable promise in protein design but still face limitations in success rate, due to the scarcity of high-quality protein datasets for supervised pretraining. We present ProteinZero, a novel framework that enables scalable, automated, and continuous self-improvement of the inverse folding model through online reinforcement learning. To achieve computationall… ▽ More Protein generative models have shown remarkable promise in protein design but still face limitations in success rate, due to the scarcity of high-quality protein datasets for supervised pretraining. We present ProteinZero, a novel framework that enables scalable, automated, and continuous self-improvement of the inverse folding model through online reinforcement learning. To achieve computationally tractable online feedback, we introduce efficient proxy reward models based on ESM-fold and a novel rapid ddG predictor that significantly accelerates evaluation speed. ProteinZero employs a general RL framework balancing multi-reward maximization, KL-divergence from a reference model, and a novel protein-embedding level diversity regularization that prevents mode collapse while promoting higher sequence diversity. Through extensive experiments, we demonstrate that ProteinZero substantially outperforms existing methods across every key metric in protein design, achieving significant improvements in structural accuracy, designability, thermodynamic stability, and sequence diversity. Most impressively, ProteinZero reduces design failure rates by approximately 36% - 48% compared to widely-used methods like ProteinMPNN, ESM-IF and InstructPLM, consistently achieving success rates exceeding 90% across diverse and complex protein folds. Notably, the entire RL run on CATH-4.3 can be done with a single 8 X GPU node in under 3 days, including reward computation. Our work establishes a new paradigm for protein design where models evolve continuously from their own generated outputs, opening new possibilities for exploring the vast protein design space. △ Less

Submitted 10 June, 2025; v1 submitted 9 June, 2025; originally announced June 2025.

Variational Supervised Contrastive Learning

Authors: Ziwen Wang, Jiajun Fan, Thao Nguyen, Heng Ji, Ge Liu

Abstract: Contrastive learning has proven to be highly efficient and adaptable in shaping representation spaces across diverse modalities by pulling similar samples together and pushing dissimilar ones apart. However, two key limitations persist: (1) Without explicit regulation of the embedding distribution, semantically related instances can inadvertently be pushed apart unless complementary signals guide… ▽ More Contrastive learning has proven to be highly efficient and adaptable in shaping representation spaces across diverse modalities by pulling similar samples together and pushing dissimilar ones apart. However, two key limitations persist: (1) Without explicit regulation of the embedding distribution, semantically related instances can inadvertently be pushed apart unless complementary signals guide pair selection, and (2) excessive reliance on large in-batch negatives and tailored augmentations hinders generalization. To address these limitations, we propose Variational Supervised Contrastive Learning (VarCon), which reformulates supervised contrastive learning as variational inference over latent class variables and maximizes a posterior-weighted evidence lower bound (ELBO) that replaces exhaustive pair-wise comparisons for efficient class-aware matching and grants fine-grained control over intra-class dispersion in the embedding space. Trained exclusively on image data, our experiments on CIFAR-10, CIFAR-100, ImageNet-100, and ImageNet-1K show that VarCon (1) achieves state-of-the-art performance for contrastive learning frameworks, reaching 79.36% Top-1 accuracy on ImageNet-1K and 78.29% on CIFAR-100 with a ResNet-50 encoder while converging in just 200 epochs; (2) yields substantially clearer decision boundaries and semantic organization in the embedding space, as evidenced by KNN classification, hierarchical clustering results, and transfer-learning assessments; and (3) demonstrates superior performance in few-shot learning than supervised baseline and superior robustness across various augmentation strategies. △ Less

Submitted 26 June, 2025; v1 submitted 9 June, 2025; originally announced June 2025.

On the Adaptive Psychological Persuasion of Large Language Models

Authors: Tianjie Ju, Yujia Chen, Hao Fei, Mong-Li Lee, Wynne Hsu, Pengzhou Cheng, Zongru Wu, Zhuosheng Zhang, Gongshen Liu

Abstract: Previous work has showcased the intriguing capabilities of Large Language Models (LLMs) in instruction-following and rhetorical fluency. However, systematic exploration of their dual capabilities to autonomously persuade and resist persuasion, particularly in contexts involving psychological rhetoric, remains unexplored. In this paper, we first evaluate four commonly adopted LLMs by tasking them t… ▽ More Previous work has showcased the intriguing capabilities of Large Language Models (LLMs) in instruction-following and rhetorical fluency. However, systematic exploration of their dual capabilities to autonomously persuade and resist persuasion, particularly in contexts involving psychological rhetoric, remains unexplored. In this paper, we first evaluate four commonly adopted LLMs by tasking them to alternately act as persuaders and listeners in adversarial dialogues. Empirical results show that persuader LLMs predominantly employ repetitive strategies, leading to low success rates. Then we introduce eleven comprehensive psychological persuasion strategies, finding that explicitly instructing LLMs to adopt specific strategies such as Fluency Effect and Repetition Effect significantly improves persuasion success rates. However, no ``one-size-fits-all'' strategy proves universally effective, with performance heavily dependent on contextual counterfactuals. Motivated by these observations, we propose an adaptive framework based on direct preference optimization that trains LLMs to autonomously select optimal strategies by leveraging persuasion results from strategy-specific responses as preference pairs. Experiments on three open-source LLMs confirm that the proposed adaptive psychological persuasion method effectively enables persuader LLMs to select optimal strategies, significantly enhancing their success rates while maintaining general capabilities. Our code is available at https://github.com/KalinaEine/PsychologicalPersuasion. △ Less

Submitted 7 June, 2025; originally announced June 2025.

Comments: Working in progress

Astra: Toward General-Purpose Mobile Robots via Hierarchical Multimodal Learning

Authors: Sheng Chen, Peiyu He, Jiaxin Hu, Ziyang Liu, Yansheng Wang, Tao Xu, Chi Zhang, Chongchong Zhang, Chao An, Shiyu Cai, Duo Cao, Kangping Chen, Shuai Chu, Tianwei Chu, Mingdi Dan, Min Du, Weiwei Fang, Pengyou Fu, Junkai Hu, Xiaowei Jiang, Zhaodi Jiang, Fuxuan Li, Jun Li, Minghui Li, Mingyao Li , et al. (46 additional authors not shown)

Abstract: Modern robot navigation systems encounter difficulties in diverse and complex indoor environments. Traditional approaches rely on multiple modules with small models or rule-based systems and thus lack adaptability to new environments. To address this, we developed Astra, a comprehensive dual-model architecture, Astra-Global and Astra-Local, for mobile robot navigation. Astra-Global, a multimodal L… ▽ More Modern robot navigation systems encounter difficulties in diverse and complex indoor environments. Traditional approaches rely on multiple modules with small models or rule-based systems and thus lack adaptability to new environments. To address this, we developed Astra, a comprehensive dual-model architecture, Astra-Global and Astra-Local, for mobile robot navigation. Astra-Global, a multimodal LLM, processes vision and language inputs to perform self and goal localization using a hybrid topological-semantic graph as the global map, and outperforms traditional visual place recognition methods. Astra-Local, a multitask network, handles local path planning and odometry estimation. Its 4D spatial-temporal encoder, trained through self-supervised learning, generates robust 4D features for downstream tasks. The planning head utilizes flow matching and a novel masked ESDF loss to minimize collision risks for generating local trajectories, and the odometry head integrates multi-sensor inputs via a transformer encoder to predict the relative pose of the robot. Deployed on real in-house mobile robots, Astra achieves high end-to-end mission success rate across diverse indoor environments. △ Less

Submitted 6 June, 2025; originally announced June 2025.

Comments: Astra Technical Report

ScaleRTL: Scaling LLMs with Reasoning Data and Test-Time Compute for Accurate RTL Code Generation

Authors: Chenhui Deng, Yun-Da Tsai, Guan-Ting Liu, Zhongzhi Yu, Haoxing Ren

Abstract: Recent advances in large language models (LLMs) have enabled near-human performance on software coding benchmarks, but their effectiveness in RTL code generation remains limited due to the scarcity of high-quality training data. While prior efforts have fine-tuned LLMs for RTL tasks, they do not fundamentally overcome the data bottleneck and lack support for test-time scaling due to their non-reas… ▽ More Recent advances in large language models (LLMs) have enabled near-human performance on software coding benchmarks, but their effectiveness in RTL code generation remains limited due to the scarcity of high-quality training data. While prior efforts have fine-tuned LLMs for RTL tasks, they do not fundamentally overcome the data bottleneck and lack support for test-time scaling due to their non-reasoning nature. In this work, we introduce ScaleRTL, the first reasoning LLM for RTL coding that scales up both high-quality reasoning data and test-time compute. Specifically, we curate a diverse set of long chain-of-thought reasoning traces averaging 56K tokens each, resulting in a dataset of 3.5B tokens that captures rich RTL knowledge. Fine-tuning a general-purpose reasoning model on this corpus yields ScaleRTL that is capable of deep RTL reasoning. Subsequently, we further enhance the performance of ScaleRTL through a novel test-time scaling strategy that extends the reasoning process via iteratively reflecting on and self-correcting previous reasoning steps. Experimental results show that ScaleRTL achieves state-of-the-art performance on VerilogEval and RTLLM, outperforming 18 competitive baselines by up to 18.4% on VerilogEval and 12.7% on RTLLM. △ Less

Submitted 15 July, 2025; v1 submitted 5 June, 2025; originally announced June 2025.

Comments: Accepted to MLCAD 2025

Evaluating Prompt-Driven Chinese Large Language Models: The Influence of Persona Assignment on Stereotypes and Safeguards

Authors: Geng Liu, Li Feng, Carlo Alberto Bono, Songbo Yang, Mengxiao Zhu, Francesco Pierri

Abstract: Recent research has highlighted that assigning specific personas to large language models (LLMs) can significantly increase harmful content generation. Yet, limited attention has been given to persona-driven toxicity in non-Western contexts, particularly in Chinese-based LLMs. In this paper, we perform a large-scale, systematic analysis of how persona assignment influences refusal behavior and res… ▽ More Recent research has highlighted that assigning specific personas to large language models (LLMs) can significantly increase harmful content generation. Yet, limited attention has been given to persona-driven toxicity in non-Western contexts, particularly in Chinese-based LLMs. In this paper, we perform a large-scale, systematic analysis of how persona assignment influences refusal behavior and response toxicity in Qwen, a widely-used Chinese language model. Utilizing fine-tuned BERT classifiers and regression analysis, our study reveals significant gender biases in refusal rates and demonstrates that certain negative personas can amplify toxicity toward Chinese social groups by up to 60-fold compared to the default model. To mitigate this toxicity, we propose an innovative multi-model feedback strategy, employing iterative interactions between Qwen and an external evaluator, which effectively reduces toxic outputs without costly model retraining. Our findings emphasize the necessity of culturally specific analyses for LLMs safety and offer a practical framework for evaluating and enhancing ethical alignment in LLM-generated content. △ Less

Submitted 5 June, 2025; originally announced June 2025.

Automated Mechanism to Support Trade Transactions in Smart Contracts with Upgrade and Repair

Authors: Christian Gang Liu, Peter Bodorik, Dawn Jutla

Abstract: In our previous research, we addressed the problem of automated transformation of models, represented using the business process model and notation (BPMN) standard, into the methods of a smart contract. The transformation supports BPMN models that contain complex multi-step activities that are supported using our concept of multi-step nested trade transactions, wherein the transactional properties… ▽ More In our previous research, we addressed the problem of automated transformation of models, represented using the business process model and notation (BPMN) standard, into the methods of a smart contract. The transformation supports BPMN models that contain complex multi-step activities that are supported using our concept of multi-step nested trade transactions, wherein the transactional properties are enforced by a mechanism generated automatically by the transformation process from a BPMN model to a smart contract. In this paper, we present a methodology for repairing a smart contract that cannot be completed due to events that were not anticipated by the developer and thus prevent the completion of the smart contract. The repair process starts with the original BPMN model fragment causing the issue, providing the modeler with the innermost transaction fragment containing the failed activity. The modeler amends the BPMN pattern on the basis of successful completion of previous activities. If repairs exceed the inner transaction's scope, they are addressed using the parent transaction's BPMN model. The amended BPMN model is then transformed into a new smart contract, ensuring consistent data and logic transitions. We previously developed a tool, called TABS+, as a proof of concept (PoC) to transform BPMN models into smart contracts for nested transactions. This paper describes the tool TABS+R, developed by extending the TABS+ tool, to allow the repair of smart contracts. △ Less

Submitted 4 June, 2025; originally announced June 2025.

Journal ref: Elsevier Journal of Blockchain: Research and Applications, 2025

Targeted Forgetting of Image Subgroups in CLIP Models

Authors: Zeliang Zhang, Gaowen Liu, Charles Fleming, Ramana Rao Kompella, Chenliang Xu

Abstract: Foundation models (FMs) such as CLIP have demonstrated impressive zero-shot performance across various tasks by leveraging large-scale, unsupervised pre-training. However, they often inherit harmful or unwanted knowledge from noisy internet-sourced datasets, compromising their reliability in real-world applications. Existing model unlearning methods either rely on access to pre-trained datasets or… ▽ More Foundation models (FMs) such as CLIP have demonstrated impressive zero-shot performance across various tasks by leveraging large-scale, unsupervised pre-training. However, they often inherit harmful or unwanted knowledge from noisy internet-sourced datasets, compromising their reliability in real-world applications. Existing model unlearning methods either rely on access to pre-trained datasets or focus on coarse-grained unlearning (e.g., entire classes), leaving a critical gap for fine-grained unlearning. In this paper, we address the challenging scenario of selectively forgetting specific portions of knowledge within a class, without access to pre-trained data, while preserving the model's overall performance. We propose a novel three-stage approach that progressively unlearns targeted knowledge while mitigating over-forgetting. It consists of (1) a forgetting stage to fine-tune the CLIP on samples to be forgotten, (2) a reminding stage to restore performance on retained samples, and (3) a restoring stage to recover zero-shot capabilities using model souping. Additionally, we introduce knowledge distillation to handle the distribution disparity between forgetting, retaining samples, and unseen pre-trained data. Extensive experiments on CIFAR-10, ImageNet-1K, and style datasets demonstrate that our approach effectively unlearns specific subgroups while maintaining strong zero-shot performance on semantically similar subgroups and other categories, significantly outperforming baseline unlearning methods, which lose effectiveness under the CLIP unlearning setting. △ Less

Submitted 3 June, 2025; originally announced June 2025.

Comments: 12 Figures,5 Pages. The project page is \url{https://zhangaipi.github.io/forget_clip/}

Transforming Automatically BPMN Models to Smart Contracts with Nested Collaborative Transactions (TABS+)

Authors: Christian Gang Liu, Peter Bodorik, Dawn Jutla

Abstract: Development of blockchain smart contracts is more difficult than mainstream software development because the underlying blockchain infrastructure poses additional complexity. To ease the developer's task of writing smart contract, as other research efforts, we also use Business Process Model and Notation BPMN modeling to describe application requirements for trade of goods and services and then tr… ▽ More Development of blockchain smart contracts is more difficult than mainstream software development because the underlying blockchain infrastructure poses additional complexity. To ease the developer's task of writing smart contract, as other research efforts, we also use Business Process Model and Notation BPMN modeling to describe application requirements for trade of goods and services and then transform automatically the BPMN model into the methods of a smart contract. In our previous research we described our approach and a tool to Transform Automatically BPMN models into Smart contracts TABS. In this paper, we describe how the TABS approach is augmented with the support for a BPMN collaborative transaction by several actors. Our approach analyzes the BPMN model to determine which patterns in the BPMN model are suitable for use as collaborative transactions. The found BPMN patterns that are suitable as transactions are shown to the developer who decides which ones should be deployed as collaborative transactions. We describe how our approach automatically transform the BPMN model into smart contract the provides a transaction mechanism to enforce the transactional properties of the nested transactions. Our approach greatly reduces the developers task as synchronization of collaborative activities is provided by our approach, so that the developer needs to code only independent tasks with well-defined inputs and outputs. We also overview the TABS+ tool we built as a proof of concept to show that our approach is feasible. Finally, we provide estimates on the cost of supporting the nested BPMN collaborative transactions. △ Less

Submitted 3 June, 2025; originally announced June 2025.

Comments: Preprint. arXiv admin note: substantial text overlap with arXiv:2505.24309

Journal ref: Distributed Ledger Technologies: Research and Practice, Volume 3, Issue 3 Article No.: 21, Pages 1 - 37, 2024

AI Debate Aids Assessment of Controversial Claims

Authors: Salman Rahman, Sheriff Issaka, Ashima Suvarna, Genglin Liu, James Shiffer, Jaeyoung Lee, Md Rizwan Parvez, Hamid Palangi, Shi Feng, Nanyun Peng, Yejin Choi, Julian Michael, Liwei Jiang, Saadia Gabriel

Abstract: As AI grows more powerful, it will increasingly shape how we understand the world. But with this influence comes the risk of amplifying misinformation and deepening social divides-especially on consequential topics like public health where factual accuracy directly impacts well-being. Scalable Oversight aims to ensure AI truthfulness by enabling humans to supervise systems that may exceed human ca… ▽ More As AI grows more powerful, it will increasingly shape how we understand the world. But with this influence comes the risk of amplifying misinformation and deepening social divides-especially on consequential topics like public health where factual accuracy directly impacts well-being. Scalable Oversight aims to ensure AI truthfulness by enabling humans to supervise systems that may exceed human capabilities--yet humans themselves hold different beliefs and biases that impair their judgment. We study whether AI debate can guide biased judges toward the truth by having two AI systems debate opposing sides of controversial COVID-19 factuality claims where people hold strong prior beliefs. We conduct two studies: one with human judges holding either mainstream or skeptical beliefs evaluating factuality claims through AI-assisted debate or consultancy protocols, and a second examining the same problem with personalized AI judges designed to mimic these different human belief systems. In our human study, we find that debate-where two AI advisor systems present opposing evidence-based arguments-consistently improves judgment accuracy and confidence calibration, outperforming consultancy with a single-advisor system by 10% overall. The improvement is most significant for judges with mainstream beliefs (+15.2% accuracy), though debate also helps skeptical judges who initially misjudge claims move toward accurate views (+4.7% accuracy). In our AI judge study, we find that AI judges with human-like personas achieve even higher accuracy (78.5%) than human judges (70.1%) and default AI judges without personas (69.8%), suggesting their potential for supervising frontier AI models. These findings highlight AI debate as a promising path toward scalable, bias-resilient oversight--leveraging both diverse human and AI judgments to move closer to truth in contested domains. △ Less

Submitted 2 June, 2025; originally announced June 2025.